Class-based outlier detection: Staying zombies or awaiting for resurrection?

Abstract

This paper addresses the task of finding outliers within each class in the context of supervised classification problems. Class-based outliers are cases that deviate too much with respect to the cases of the same class. We introduce a novel method for outlier detection in labelled data based on Random Forests and compare it with existing methods both on artificial and real-world data. We show that it is competitive with the existing methods and sometimes gives more intuitive results. We also provide an overview for outlier detection in labelled data. The main contribution are two methods for class-based outlier description and interpretation.